Locomotor strategies inside the blind walks: freely moving previsual rat pups in the open field test

Abstract

Blind walks of previsual rat pups in the open field test were considered as a supposedly random process. At this age, immature rats should rely only on non-visual modalities for sampling a novel space and constructing putative cognitive maps. Spontaneous locomotion of outbred Wistar rat infants (n=51) on their 13th postnatal day was tracked for 2 minutes and analyzed offline for the presence of random components and/or strategies of locomotion. Three distinct patterns were observed. A large portion of the pups (n=22) performed sub-diffusive localized walks in the center. A smaller cohort (n=9) undertook almost immediate quasi-linear raids in a random direction, thus succeeding in reaching a shelter (i.e., a wall vicinity). The rest (n=20) demonstrated a mixed strategy: localized walks interspersed with quasi-linear raids. An algorithm for automated segmentation of the open field trajectories into the localized walks and quasi-linear fragments was developed; the fragments were analyzed separately. Statistical analysis shows the localized walks to be diffusive (i.e., Brownian) only within 2-3 seconds, but essentially sub-diffusive for longer time scales. The autocorrelation function shows that sub-diffusion was caused by the re-attraction of trajectories. Self-odor traces can be a physical cue ensuring the effect. Gender or body weights were not significant predictors for any locomotor parameters. The localized walks can be considered as a primary level of idiothetically-cued path integration in previsual pups, whereas the run trials can be a form of an active escape response. The applied method of space potentials avails the automated functional segmentation of locomotor tracks and thus brings advantages of big data analysis for studying non-visual navigation in animals.